CHAPTER 5

Geodynamic models of seismically active zones

The information provided in the previous chapter shows that within the Arctic Region both existing types of seismically active zones are presented: interplate and intraplate. This fact has been generally established long ago, however a clear formulation was first given in the review by (The Arctic Ocean Region..,1990). Virtually, a single and the only interplate seismicity zone is the Mid-Arctic Earthquake Belt consisting of aforesaid numerous, genetically related fragments and crossing the region approximately into two equal parts. Almost all remaining can be reasonably assigned to intraplate zones which are not directly related to any global seismic belt.

The objective of this chapter is to generalize current ideas on the geodynamics of seismically active zones using the most recent seismological material; to study geodynamics of some areas in interplate zones which are presently less known; and to critically evaluate the tectonic nature of intraplate seismicity.

5.1.Interplate seismicity zone (The Mid-Arctic Earthquake Belt)

At present, it is found to be definite that modern tectonics of the most of the Arctic Region is subject to interaction of two major lithospheric plates: Eurasian and North American. The Mid-Arctic Earthquake Belt marks the oceanic part of the modern plate boundary passing along the subsea Kolbeinsey, Mohns, Knipovich ridges and dividing them Tjornes, Jan Mayen, Spitsbergen fracture zones in the Norwegian-Greenland Basin, and the Gakkel Ridge up to the Laptev Sea continental slope in the Eurasian Subbasin (Figs. 6,8,11).

Basic geological and geophysical materials allowing recovery of the history of emergence and evolution of the above boundary and genetically related geomorphological and tectonic structures it crosses, as well as their modern evolutionary trends, are magnetic and seismological data. According to large-scale airborne, marine and ice national and foreign magnetic surveys within the Norwegian-Greenland Basin and Eurasian Subbasin, a band-shaped magnetic field is established to be symmetrical with respect to the axial line of the ridge. This field is typical of oceanic basins formed as a result of the lithosphere split and subsequent ocean floor spreading. Along the entire length of the basins, positive anomalies of the same name are traced with various extent of reliability. The oldest of them are anomalies of the age 24 (56 million years - the Lower Eocene) fringed along the basinal periphery by a band of the negative magnetic field. In the North Atlantic and Labrador Sea the oldest are the anomalies of the age 34 (70 million years - the Late Cretaceous). In the Labrador Sea anomalies younger the age 13 (36 million years - the Early Oligocene) have not been identified. Based on analysis of both band-shaped magnetic anomalies which are isochrones of opening, and fracture zones showing the direction of opening, it has been established that between anomalies 34 and 25 (the Late Cretaceous - Late Paleocene) spreading existed in the North Atlantic, Labrador Sea and Baffin Bay. During anomaly 25 the lithosphere split in the Norwegian-Greenland Basin and Eurasian Subbasin, and beginning the Early Eocene (anomaly 24) active spreading started in this area. Triple junctions north and south of Greenland were formed, and the Greenland Plate which thereby emerged shifted with respect to both North American and Eurasian plates. In the Eurasian Subbasin the Lomonosov Ridge started to get separated from the modern Barents-Kara shelf. This geodynamic environment continued until the Early Oligocene (36 million years - anomaly 13) when spreading in the Labrador Sea and Baffin Bay stopped, triple junctions disappeared, and the Greenland plate became part of the North American plate.

Conclusions drawn from magnetic data regarding a divergent nature of the North American and Eurasian boundaries in the Norwegian-Greenland Basin and Eurasian Subbasin are confidently supported by seismological data elucidating the current geodynamic environment in the region.

It is clear that the type of the boundary, its spatial location, are defined from co-effect of two principal factors: orientation of tectonic forces and real characteristics of the initial lithosphere. The effect of the second factor is responsible for each particular fragment of the boundary, and accordingly, for the parameters marking this boundary of the seismic belt. In the subuniform and subisotrope lithosphere the orientation of the eventual rupture plane would be primarily defined by the laws of mechanics and would depend on the orientation of the forces applied. If a pre-spreading weakened zone existed in the initial lithosphere, the emerging rupturing line would rotate from the theoretical position towards the weakened zone. The larger angle between the weakened zone strike and approaching spreading split, the higher rotation. In some particular cases, when the angle is equal to zero, we obtain in fact the aforesaid type of the split in the subuniform lithosphere. In case of right or nearly right angle, for example, when the approaching split directly abuts a rigid monolith block, we encounter the type where the rupturing line along the nearly orthogonal fault or series of en-echelon faults bends around the block. Therefore three basic types exist:

- absence of pre-spreading weakened zone or coincidence of its strike with that of the spreading split;

- strike of the weakened zone and spreading split form an acute angle;

- weakened zone and spreading split form a right or nearly right angle.

A certain distribution of earthquake epicenters and a type of focal mechanism should correspond to each of these three cases. Factual data on the Mid-Arctic Seismic Belt show that all three aforesaid versions are presented here.

The Mohns Ridge and Gakkel Ridge correspond to the first type. Linearity of these ridges along hundreds and even thousands kilometers, their subparallel position in respect to each other and to the averaged axial line of the belt passing across the rotation pole in the northern Yakutia, the obvious predominating role of the normal-fault focal mechanism and mutual orthogonality of the extension stresses and axis of the ridges (Figs 7, 9, 12), narrow linear distribution of the epicenters near the crest of the ridge is, in our view, a convincing evidence that rift processes here began either in the conditions of fairly uniform lithosphere or along the fault concordant to the extension, and follow a scheme close to classic. This is confirmed by geomorphological characteristics of ridges as well: their narrowness (not more than 200-250 km), drastic rising over adjacent abyssal (500-1500 m), and existence along nearly the entire length of the ridges of a narrow (1-5 to 10-30 km) and deep rift valley. In view of the fact that rifting split as a rule "searches for" weakened zones, the type of pre-rifting concordant weakened zone seems to be preferable. As to the Eurasian Subbasin, V.I.Ustritsky (1990) shares this opinion suggesting the rift emplacement along the weakened zone formed and revived from time to time during the Phanerozoic.

The above regularity of geomorphological and seismological characteristics is most heavily disturbed only at the site of the Gakkel Ridge between 35° E and 80° E. Here, almost linear fragment of the seismic belt about 300 km long is displaced 100-120 km northward in its western part, and farther east, it fairly smoothly approaches the general axial line. In the zone of rupture and belt displacement a strike-slip mechanism typicalof transform faults is established. According to bathymetry data, the rift zone at this site becomes less distinct and locally, completely disappears, while the ridge itself is morphologically more obscure. Interestingly, the boundaries of the above site of disturbed linearity of the ridge and seismic belt are in range line with identified on the shelf submeridional troughs Franz-Victoria, Saint Anna and Voronin. At intersection of these with the continental slope some earthquakes with magnitude up to 5.5 occur. This suggests pre-rifting existence of these troughs which played the part of weakened zones at this site. These weakened zones distorted the classic rift development. It should be noted that according to Ustritskiy¸s (1990) compilations, in this area, at least beginning the Triassic, a boundary between the zones with essentially different paleogeographic environments is marked.

The third type of the boundary modally opposite to the first one is presented by the Tjornes, Jan Mayen and Spitsbergen fracture zones where the plate boundary (the axis of the ridge) is displaced to either side dozens or even hundreds kilometers off. For the first time, geodynamics of such zones was studied by J.Wilson (1965) who identified the transform fault model accounting for the sinistral strike-slip mechanism at dextral displacement of the boundary and vice versa, and existence of seismicity only at the sites of the zones restricted by displaced boundary fragments.

Most data on focal mechanisms support the transform nature of the aforesaid zones: subvertical nodal planes, one of which coincides with the fractures zone, and predominance of the strike-slip mechanism. In the Tjornes and Spitsbergen fracture zones where the axis of the ridge is displaced to the left, a dextral strike-slip is confidently identified while a sinistral strike-slip is established in the Jan Mayen zone where displacement is to the right.

One should keep in mind, yet, that in reality no environment can be perfectly uniform. Therefore, identification of one or another fragment of the boundary with one of the two aforesaid types is tentative, and depends on the approximation degree of the environment. An illustration to this statement are the above mentioned parts of disturbed linearity of the Mid-Arctic Belt within the Gakkel Ridge zone corresponding to the first type.

The second, intermediary type of boundary where so called "oblique" extension can be expected is represented by the Knipovich Ridge. W.Harland (1979) was the first to look at the geometry of movements within such zones. He identified two possible types of mobile belts (plate boundaries), general strike of which is obliquely oriented towards the direction of the extension. One of them, stepwise in the plane, is in fact a combination of the first and third types of the boundary, i.e. a system of mutually normal alternating sites of direct extension and transform faults. The strike of these faults is under an angle towards the general strike. In the other, there are no fragments with a strike differing from the general, and here it is a single persistent line. It is evident that the crucial part in the formation of one or another form of the split is played by the inner structure of the pre-spreading lithosphere.

A typical example of the stepwise boundary is the area of the Mid-Arctic Ridge near the equator. In case of the Knipovich Ridge, the situation is much less clear due to its two main features. First, near the equator, where despite radical changes in the strike, the ridge along the entire length covers the median zone of the basin. The Knipovich Ridge as it has been mentioned above, is drastically displaced eastward. The width of the oceanic basin between the Knipovich Ridge and the Spitsbergen continental slope does not exceed 100 km narrowing northward down to 20-30 km. The western flank is fringed with a deep sea basin which is several times wider. Second, anomaly magnetic fields above these fragments of the mid-oceanic ridge are evidently different. Near the equator a typical spreading field has been established (Geophysical characteristics..,1985, etc.) with rather confidently identified pairs of submeridionally oriented magnetic anomalies. Fragments of these anomalies in association with displacing transform faults form a number of stepwise lines. The main peculiarity of the anomalous magnetic field over the Knipovich Ridge is its mosaic pattern and fairly weak manifestation of features typical of spreading basins (Geophysical characteristics..,1985; Johnson and Heezen,1967). The aforesaid distinctive features of the Knipovich Ridge did not prevent the group of Russian scientists leaded by A.M.Karasik from developing an elegant concept presenting the Knipovich Ridge as a typical mid-oceanic ridge (Geophysical characteristics..,1985; Desimon and Karasik,1979, etc.). Based on the analysis of the magnetic field they draw a conclusion that its mosaic pattern is due to the fragmentation of linear anomalies by a dense system of transform faults. These faults displace fragments of the anomalies westward. An illusion of oblique spreading is caused by non-coincidence in the general (integral) and real (fragments) strike of the axes of the anomalies. The former is close to the meridional, sharply different from that of the Mohns Ridge. The latter shows azimuth of 30-40° which is much closer to that of the ridge and accordingly, spreading along the azimuth of 120-130° . Therefore, based on the data of the above scientists, the area of the Knipovich Ridge as per dynamics and kinematics of the lithospheric evolution shall be recognized as similar to the area of the Mid-Atlantic Ridge near the equator. Spreading along the azimuth of 120-130° ensures the involvement in the area of the spreading oceanic lithosphere east of the Knipovich Ridge of a much wider band where magnetic anomalies up to 13 (38 million years) inclusive may be well placed. This was a partial solution of a problem regarding asymmetrical location of the ridge in the basin. For complete solution, an assumption was made on spreading axis jumping during the period between anomalies 13 and 14. A precedent is known from the southern part of Norwegian-Greenland Basin. Most of scientists dealing with the problem recognize the spreading axis jumping between anomalies 20 and 23 from the presently aseismic Aegir Ridge (Fig.7) into the modern position along the axis of the Kolbeinsey Ridge.

Paying tribute to the harmony and attractiveness of this concept we consider it necessary to state that it is seems to us fairly artificial and inconsistent with the reality. Without giving details of the incredible complexity of identification and recognition of coupled anomalies in the conditions of the mosaic magnetic field on the Knipovich Ridge and surrounding offshore areas, we state two points falling out of the aforesaid concept It is evident that the nature of the magnetic field and geomorphology of the Ridge should be closely correlated with each. Fragments of linear anomalies displaced parallel to themselves along transform faults and showing a strike different from the general one should becorrespondent to similar fragments of the Ridge itself. Nothing like is identified in reality. The Ridge, although split by a system of transverse faults, is presented by a series of narrow subparallel uplifts conjugated with narrow trenches suffering no heavy transverse displacements.

Another point is that in contrast to the southern part of the Norwegian-Greenland Basin where the site of ancient spreading axis may be identified with the Aegir Ridge, west of the Knipovich Ridge neither bottom topography nor the basement relief nor geophysical fields show indications of the existence of anything like that.

In our view, particular features of the modern geomorphology of the Knipovich Ridge, its fault tectonics, seismicity and dynamics are defined by the fact that the interplate split propagating from the south has penetrated under an non-right angle into the confines of complex arrangement of the lithospheric block similar to the modern Spitsbergen block which had previously been its western part. The frontal boundary of this block was marked along the fracture zone presumably colinear with the modern Senja Fracture Zone. This line was of the same north-northwesterly strike as the main Caledonian fault system on the Svalbard Archipelago (Billefjorden, Lomfjorden, Hinlopenstretet, etc.) (Fig.9), and also the Hornsund Fault located west of Svalbard in the zone of the continental slope. Similar strike has been inherited by mountainous ranges and depressions of the Knipovich Ridge. The faults crossing the Ridge are also connected with the fault system of appropriate strike. These strikes on Svalbard had secondary implication during the Caledonian folding. However, they manifested only during the Cenozoic tectonic reanimation resulting in the formation of major latitudinally oriented fjords (Isfjorden, Bellsund, etc.). A major transverse zonality of the Knipovich Ridge is clearly manifested in the distribution of earthquake epicenters forming three clusters divided by zones with drastically reduced seismicity (Fig.7). It is easy to notice that the similarly oriented major block zonality is easily found on West Spitsbergen Island. It is striking that sites with lower seismicity on the Knipovich Ridge are located in a range line with the strike of submerged blocks:Sorkapp Trough south of the island, the system of strong Isfjorden and Belsund in the center, and marine fringing of the island in the north. At the transect of the most distinct clusters of epicenters near 76° and 78 ° N there are uplifted blocks of the southern and northern parts of the island. The above regularity is undoubtedly unsuitable for making analogy of modern geodynamics in the Knipovich Ridge and Svalbard Archipelago areas. It just confirms our view that both aforesaid regions were an integrity during the pre-spreading epoch. Accepting this view one can make the following conclusion. Superimposition on extending riftogenic forces obliquely oriented with respect to pre-spreading dislocations within the block of the lithosphere including the site of the contemporary Knipovich Ridge, has lead to the emergence of tectonic regime with elements of both riftogenic and transform. In contrast to near equator area of the mid-Atlantic Ridge, where the combination of transform and rifting mode takes place, but their activity is laterally spread (partial alternation of transform faults and displaced fragments of the ridge), in the area of the Knipovich Ridge each element of the lithosphere is involved in the complex movement. The orientation of this movement is defined by the direction of rifting forces (northwest-southeast ) and strike slip along the faults of submeridional strike. The part played by the strike slip seems to be greater. This is attested by data on focal mechanisms and by greater proximity of the strike of post-spreading Knipovich Ridge not to that of Mohn Ridge but to the strike of major faults of Spitsbergen. This is a good explanation to the existence of a fairly narrow band of the oceanic lithosphere between the Knipovich Ridge and Svalbard. Conjugated under an obtuse angle with the Knipovich Ridge the Spitsbergen Fracture Zone which has inherited its position from an older weakened zone, has a strike coincident with the orientation of extending forces and is typical transform fault where the extending component is almost equal to zero.

Therefore, the system "Knipovich Ridge - Spitsbergen Fracture Zone" represents a complex link between the two laterally displaced by almost a thousand kilometers elements of a single system of the mid-oceanic Mohns and Gakkel ridges. Particular features of this system are crucially defined by those of the pre-rifting structure of the lithosphere which ensure the existence in it of two conjugated elements: one with reduced and the other with zero divergent component.

In terms of the statement of a character of the newly forming split superimposed on the pre-rifting lithospheric structure, available seismological data make it possible to assert that with transition across the continental slope to the shelf of the Laptev Sea a weakened zone was found not on the North graben as assumed earlier, but eastward in the area of the folded mesozoid complexes of the Northeastern Eurasia. There the epicentral line extended abutting the junction of the shelf and the block of the New Siberian Islands immediately west of Belkovskiy Island. It should be noted that it was first noticed by B.I.Kim (1986). Scattering of epicenters farther south is an evidence of the transformation of one principal weakened zone into a series of less evident, which cannot be considered unexpected in the folded area. To the south of Stolbovoy Island, where seismically active zone is actually degenerated, higher monolithic lithosphere and impossible further propagation of the split in this area may be suggested. Continuing activity of riftogenic forces should have lead and did lead to the occurrence of weakened zones at other sites. The position of these ancient weakened zones at present is elucidated in our view, by the entire branched system of the above troughs within which strong earthquakes took place. The most distinctly it is seen with reference to the Ust- Lena Trough which had been at a certain stage of its evolution a riftogenic extension axis. This is evidenced, in particular, by the establishment in the zone of the Buor-Khaya Gulf according to deep seismic sounding (Kogan,1974) and seismological data (Avetisov and Guseva,1991) the reduction of seismic wave velocity in the upper mantle. However, the above information on the predominance in the southern part of the trough of horizontal transverse compression shows that presently the extension axis does not pass here. In our view, seismological data make us admit that it is presently located in the area of junction of periphery flexural-fault limits of the southwestern part of the Laptev platform and Lena-Taymyr zone of conterminous uplifts. The jump of the axis seems to be genetically related with the movement from the south towards the Buor-Khaya Gulf of the North American and Eurasian plate rotation pole that took place some 1-3 million years ago (Cook and others,1984;1986). At the same time in the Moma Graben the riftogenic regime changed for the transverse compression regime which has been recently established on the basis of seismological data. The above seismological data which are at present the most complete, show that the modern understanding of the persistence on the Laptev Sea shelf of the Eurasian and North American plates boundary (Grachev,1977; Grachev,1982; Grachev and others,1967; Gramberg and others,1990; Karasik,1968, etc.) may be adopted in the first approximation only, i.e. if the entire shelf is considered to be a part of the boundary. It is sufficient to understand the general kinematics of the plate tectonic motions, however does not satisfy regional research requirements. It is necessary to recognize that there are two "blind" segments of this boundary, one of which in the eastern half of the shelf is the termination of the oceanic part of the boundary extended from the Eurasian Subbasin, and the second is that of the continental part passing from the eastern Yakutia. Proximity of the sites with predominant horizontal extension and compression suggests a violation of the postulate of "rigidity" of lithospheric plates confirming our previous statement on the existence of induced compression areas on the flanks of extension zones (Avetisov,1975; 1979, etc.). This is supported by geological data on the New Siberian Islands (Savostin and Drachev,1988).

Characteristic fact is the nearly oval-shaped southern termination of the Eurasian Subbasin and its great width (500-600 km) in the junction with the Laptev Sea continental slope. According to A.M.Karasik (1968, etc.) who had traced almost up to the slope the entire range of spreading magnetic anomalies from the oldest

(24) to the youngest (5) one, the width of the spreading zone is commensurable with the width of the Subbasin. In this case it s fairly hard to explain the drastic disappearance of such a wide spreading zone in the course of transition over the continental slope. The first thought on the transverse displacement of the "coasts" of the Eurasian Subbasin, particularly, the Lomonosov Ridge, with respect to the shelf should be abandoned. It is evident that such great-scaled motions would not have occurred without seismic manifestations. These could not have been fallen out of sight even taking into account the long distance between the region and recording stations. It remains only to suggest that the lithosphere of the Laptev Sea shelf enjoys specific plasticity allowing the formation of a series of local faults only and preventing form their integration into a single split. It is quite possible that anomalous elasticity of the lithosphere in this region is the cause of frequent oscillations in the motion kinematics identified from multiple meridional displacements of the spreading pole.

Another interpretation which does not require assumption on the anomalousness of the elastic properties of the Laptev Sea shelf lithosphere may be offered, if the magnetic anomaly axis map presented by north American scientists (The Arctic Ocean Region..,1990) is taken as a reference point. According to their data, southward, older anomalies gradually disappear, and the foot of the continental slope is reached only by the youngest one. The distance between the pair of axes in this anomaly (50-60 km) is the measure to evaluate the width of the extension zone in the southern termination of the Eurasian Subbasin. To explain the great width of the Subbasin it is necessary to assume here its pre-rifting existence, that is no junction of the Lomonosov Ridge with the Severnaya Zemlya continental slope. This assumption as just one explication of 50-100 km wide band of the negative magnetic field outside anomaly , has been previously made by P. Vogt et al. (1979). Ancient Mesozoic basins likely controlled the position of the initial line of the modern spreading split. It should be noted that based on the analysis of the topography and structure basements of the Norwegian-Greenland Basin and Eurasian Subbasin, Senin and others (1989) have come to a conclusion on the existence of pre-spreading Mesozoic basins within their confines. These were responsible to a large extent for the position of the axial line of the modern spreading.

The above data on the Laptev Sea shelf, in our view, make it possible to suggest that the formation here of a unified boundary of the Eurasian and North American plates is controlled by two counter movements of its broken fragments: to the south through the mesozoid area in the eastern part of the shelf, and to the northwest along the fault limits of the Lena-Taymyr zone of conterminous uplifts. At present, one can make only general assumption regarding the character and location of the possible connection of broken fragments in the case of continuing activity of riftogenic extending forces. It is evident, that the expected position of the unified boundary will be again to a large extent defined by the position of weakened zones existing in the region. The following possible variants seem realistic.

1. Southward advance of the eastern split fragment along the Omoloy and southern Ust- Lena grabens and connection with the continental part of the boundary east of the Lena delta; north-northwestward advance of the western fragment along the contact of the Taymyr folded system with the western part of the Laptev platform and its exit to the continental slope. This would result in the formation of the Laptev Sea microplate and triple junction near the south coast of the Buor-Khaya Gulf. Within this solution an exit of the western split fragment to the continental slope is possible through the North Graben.

2. Stop in the advance of splits in the previously mentioned direction and connection of the presently identified ends of the boundary fragments along the northeastern strike line coincident or close to the North Graben. In this case, formation on the Laptev shelf of a transform fault system like Spitsbergen should be expected.

For gaining a better understanding of the modern dynamics of the lithosphere in the Mid-Arctic Earthquake Belt zone it is interesting to compare the levels of seismic activity of its individual fragments. It should be noted that the first attempt was made by R.M.Demenitskaya and E.M.Litvinov (1974). One should note, though not quite unequivocal but admissible in this case, logical chain:

- in the movement of rigid plates on the sphere the spreading rate is proportional to the latitude of the site with respect of the rotation pole;

- taking into account the modern position of the rotation pole of the Eurasian and North American plates which despite a great discrepancy in definitions (Chapman and Solomon,1976; Cook and others,1986; Karasik and others,1975; Pitman and Talwani,1972; Savostin and Karasik,1981, etc.) may be recognized as located within the Northeastern Asia, the spreading rate should increase from the Laptev Sea shelf towards the Atlantic;

- since the intensity of tectonic movements directly affect the level of seismic activity the latter, in general, should regularly increase in the same direction, as the spreading rate. Ambiguity of this statement is primarily that it is effective only in a restricted range of rates. No highest seismicity is noted in the superfast spreading zones (e.g. East Pacific Rise - 15 cm/yr). This probably related to the fact that in these cases the crucial part is played by increased plasticity of the lithosphere.

Available definitions of modern spreading rates, that is increase by 0.3 to 0.7 mm/yr from the east to the west in the Eurasian Subbasin on the Gakkel Ridge, and by 12-15 mm/yr in the Norwegian-Greenland Basin on the Mohns Ridge (Karasik and others,1975, etc.). in general, fall within the scope of theoretical expectations. On the Knipovich Ridge the same authors basing on the postulate of plate "rigidity" forecast intermediate values of 9-10 mm/yr.

Estimation of the general level of seismic activity in certain parts of the Mid-Arctic Seismic Belt has been made by means of determination over the entire period of instrumental observations of the total amount of earthquakes with magnitude 5 (completeness threshold) and more and reduction of this amount to the length unit of 1000 km.

Table 42 shows that even this simplified comparison provides a useful information which partially turns out to be anticipated but partially quite unexpected. The former is a high level of activity of the Mohns and Kolbeinsey ridges in comparison with the Gakkel Ridge, and maximum levels of activity of the Spitsbergen and Jan Mayen fracture zones. The latter conforms to the actual duplication of the velocity with reference to the movement of the transform fault walls since they belong to different plates. Unexpected is undoubtedly a low activity of the Knipovich Ridge. This fact, firstly, challenges the whole scheme according to which the Knipovich Ridge is crossed by a dense fault system. Secondly, it becomes evident, that the plates, at least in this area, do not behave as absolutely rigid ones. A reduced seismicity on the Knipovich Ridge may be attributed to the fact that stress discharge takes place not only on the interplate boundary. This conclusion splendidly conforms to the high seismicity of the Senja Fracture Zone and Lofoten Basin located actually in the southern range line with the Knipovich Ridge, and seismicity of the Svalbard Archipelago. Some fairly strong earthquakes occur within the oceanic part of the Greenland Sea.

Table 42

Comparison between seismic activity levels
of various segments of the Mid-Arctic Earthquake Belt

Not less unexpected turned out to be the low activity of the Reykjanes Ridge. On either side of it epicenters of fairly strong earthquakes have been reported from oceanic basins.

Therefore, based on the all available integrated geological and geophysical information, including that presented above, it should be admitted that the Transarctic boundary of the Eurasian and North American plates persistent in the plane and evolving in general under a single scheme, yet consists of a series of mutually connected segments evolution of which, when scrutinized, has specific features of their own. These features are defined by changing properties of the actual geological environment and, primarily, by the fact that it should be regarded absolutely rigid only to a certain extent of approximation.

5.2. Intraplate seismicity zones

The previous chapter has shown that seismicity not related to the interplate boundaries is fairly well presented in the Arctic Region. The common view presently adopted suggests that it emerged here due to the different effect of three factors: partial discharge of stresses generated in the Transarctic interplate zone, ancient glacial rebound and lithosphere response to the thick overburden. It is quite evident that the above factors have different extent of influence, thus the contribution of any of them to the total effect much depends on the location and evolutionary history of each particular region. The first factor has transarctic influence, and it is natural to assume that its contribution to seismicity should be proportional to the distance between the region and the closest interplate zone. The other two factors are local, and their influence zones can be easily predicted and limited.

Although the fact of occurrence of higher seismicity is due solely to external geological and tectonic forces resulting in the creation in the lithosphere of excessive stresses, specific features of seismicity, including earthquakes intensity, hypocenters, focal mechanisms are equally dependent, as it has been mentioned above, on the peculiar features of the real geological environment affected by these forces. In the areas of interplate seismicity the intensity of outer tectonic forces is sufficient to transform the lithosphere, and the part of the initial structure is in the adjustment of the location and form of the new structural tectonic elements. In the zones of intraplate seismicity this part is much more significant as the new structural pattern is, in fact, the revived old one. As a consequence which has both scientific and applied significance, particularly in seismic hazard assessment, may be a conclusion that earthquakes most frequently first occur in previously existing weakened zones of the lithosphere. That is why when presenting geodynamics of intraplate seismic zones and estimation of tectonic nature of intraplate earthquakes much more attention than for interplate seismicity should be paid to information on structural tectonic features of the regions.

When looking at intraplate seismicity zones in view of their scattering it seems useful to take advantage of the indication of geographic location and subdivide them into three major groups: zones related to the continental and shelf (insular) fringing of the Norwegian-Greenland Basin and Eurasian Subbasin, Arctic Canada zone and adjacent regions, and near- Pacific region.

Continental margins of the Norwegian-Greenland Basin and Eurasian Subbasin

Ideas on partial discharge in intraplate regions of stresses generated on the plate boundaries initially based on the general logic, has been presently supported by series of factual data, especially on Fennoscandia. Research into this problem is carried in two directions.

First, regional stress fields observed by various methodsare compared with expected ones. The position of a region in question relative to the nearest segment of interplate boundary and specific fault tectonics of the region is taken into account. According to the numerous data on direct measurements in wells, mining sites and mines, a bulk fault plane solutions of weak earthquakes and geological investigations of fault zones is unequivocally established that the main feature of the regional stress field of Fennoscandia is the existence of a horizontal or subhorizontal compressive component of predominant NW-SE orientation. This is in compliance with the lithospheric dynamics of Mohns Ridge, the nearest segment of the mid-oceanic ridge (Bungum,1989; Claub and others,1989; Hast,1958; Kvamme and Hansen,1989; Slunga,1989; Stephansson and others,1987; Turchaninov and Markov,1966; Wahlstrom,1989, etc.). Meanwhile, the extent of stability of the general solution expressed in scattering of singular determinations with respect to an average one is definitely related to the orientation of fault systems in different areas of the region. As it has been logically inferred and presently supported by observations the extension in axial zones of basins should have resulted and results in the formation of primarily shear stresses in weakened sites of continental margins orthogonal to basinal axes. With orientation of the weakened zone changed compression component is increased. Apart from Fennoscandia, it is well traceable along the Eurasian margin of the Norwegian-Greenland Basin and Eurasian Subbasin where earthquakes tend to transverse faults and trenches. Thus, epicentral clusters are reported from the vicinity of the Senja Fracture Zone in the Lofoten Basin and from the Franz-Victoria and Voronin troughs. Also seismicity is known from the Saint Anna Trough. (Avetisov and Golubkov,1971; Avetisov,1971). The data available here on focal mechanisms evidence for strike-slip and thrust - strike-slip regimes, one of subvertical nodal planes being close to the fault plane. Predominance of horizontal compression is also reported from backside parts of continental margins (Kola Peninsula, Novaya Zemlya) (Assinovskaya,1991; Assinovskaya and Soloviev,1993; Assinovskaya,1994), compression areas on flanks of extension zones are reported from the Laptev Sea (Avetisov,1975; 1993a), the New Siberian Islands (Savostin and Drachev1988).

Proximity of Svalbard to the Mid-Arctic Belt also allows its seismicity to be connected with the discharge of stresses generated in the area of oceanic spreading. Lack of uninterrupted zones of seismic activity connected by the Mid-Arctic Belt has lead L.Sykes and Sbar to this conclusion in 1973 (1973). In these terms, new seismological data collected since that time are well explicable (Bungum and others,1982; Chan and Mitchell,1985; Mitchell and others, 1990). On West Spitsbergen Island epicenters of earthquakes in the zone of intersection between the sublatitudinal De Geer Fault and the fault of the NNW system dominating on the Archipelago rather tend to the former which is orthogonal to the nearest fragment of the mid-oceanic ridge - the Knipovich Ridge (Fig.9). At the same time, in the north, Nordaustlandet earthquakes are grouped along the dominating fault system suborthogonal to the nearest in this area Gakkel Ridge. Fault plane solutions have provided a strike-slip mechanism. To the south on the same island, in the latitudinal-strike epicentral cluster a thrust mechanism caused by subhorizontal meridionally oriented compression has been established. It is hard to agree with the view expressed by Mitchell and others (1990) that local seismicity on Svalbard is caused by the existence in fissures and microfissures of weakened zones of liquid inclusions reducing the environmental strength. In this case higher seismicity should be anticipated most likely on the northwestern West Spitsbergen near the Bockfjorden fault where Quaternary volcanoes and related thermal springs are known.

In our view, fairly interesting is the fact of association of higher seismicity zones to the sites of lower heat flow, which has been found by the Kola geophysicists within the Barents-White Sea region (Tsibulya and others,1993). It is evident that this may only be attributed to induced or as it has been mentioned above (Avetisov,1975), "passive" character of seismicity caused by release of stresses generated outside the region. The coldest blocks of lithosphere are the most susceptible to them due to their higher brittleness.

So far there have not been data confirming compliance with interplate stresses for the Greenland margin of the Norwegian-Greenland Basin. The character of stresses may be judged here only from three fault plane solutions of the earthquakes that occurred in the offshore area nearby the shoreline (Fig.8, Tables 25, 25a). All the three solutions, yet one being poorly constrained, have provided a normal fault or close to it mechanism with the cross shoreline extension axis. It should be noted that the same distribution of stresses evidencing for the bottom subsidence and land uplift takes place in the eastern offshore zone of the Baffin Island.

The second direction of investigations into the influence of tectonic processes in the axial spreading zones on the formation of seismogenic stresses of marginal zones is comparison of seismicity regimes in these areas. Unfortunately, these investigations based on statistic data, are limited in application due to an obvious fact that their results may be only representative for areas of fairly high seismicity where data of long-term (at least a few dozens years ) reliable seismic monitoring are available. At present, in the region under consideration, only data on Fennoscandia can meet these requirements, but even here instrumental data of the first half of the century, especially of the war years are unlikely to be representative. Yet, E.Skordas et al. (1991) have carried out this work. They calculated and plotted the seismic energy released in the axial zone of the Norwegian-Greenland basin and in Fennoscandia over 70 years beginning in 1917. An evident similarity of both plots has been established (mutual correlation coefficient reaching 0.7-0.8), with temporal delay for Fennoscandia varying from 0 to 3-4 years.

Acknowledging unconditional influence of interplate geotectonic processes on the formation of intraplate seismicity one should admit the striking non-similarity in the manifestation of the influence. Taking into account similar half-width (distance between the axial zone and the margin) of the Norwegian-Greenland Basin and Eurasian Subbasin, one should expect, even in view of environmental inhomogeinity, approximately equal average level of intraplate seismicity along the periphery of the basins. In fact, as Fig.6 shows, the distribution of intraplate seismicity is spot-like. Primarily, noteworthy is definitely lower level of seismicity on the margins of the Eurasian Subbasin despite of its less width in comparison with the Norwegian-Greenland Basin, and higher seismicity of Fennoscandia in comparison with adjacent areas. In our view this may be explained by three major reasons.

The first one may be referred to as a subjective and is related to the aforesaid irregularity of the recording stations network. In Fennoscandia dozens of stations are now operating, while on a vast area between the Kola and Chukchi peninsulas their number did not exceed 5-7 in different years. At present, there is none between the stations of Apatity and Iultin. Incompleteness of our knowledge regarding seismicity of the Eurasian Subbasin and adjacent areas is convincingly proven by field seismological observations carried out by NIIGA-NPO "Sevmorgeo" (Avetisov,1975, etc.), which showed the existence of previously unknown zones of higher seismicity, particularly in the area of the New Siberian Islands. Earthquakes reported in the course of these observations north of the New Siberian Islands suggest that there is a certain seismicity level on the Lomonosov Ridge different from zero, which is also supported by recording of earthquakes at its junction with the continental slope of the Northern Greenland and Canadian Arctic Archipelago.

The second possible reason, in our view, is the lower spreading rate along the subbasinal axis due to its proximity to the rotation pole. It may be assumed that as a result of slow accumulation of stresses their considerable part is discharged due to aseismic motions in weakened zones.

Finally, the third and, likely, the principal reason is that the zones of higher seismicity emerge primarily where regional stresses transferred from axial basinal zones are complemented by the effect of the other, probably even stronger in this region, sources of excessive stresses.

This statement is primarily justified with reference to Fennoscandia - the most seismically active area within the study region (Fig.22). Both its major component such as the Baltic Shield, and related Norwegian Caledonids presently experience intensive uplift distinctly manifesting even in the historical memory of the mankind. Maximum cumulative amplitudes of uplift up to 250 m over the late postglacial period take place in the central and northern part of the Gulf of Bothnia in the Baltic Sea (Fig.23). In respect of Fennoscandia, based on the data on focal mechanisms evidencing, as it has been mentioned above, for predominance of strike-slip movements along subvertical planes over it largest part the prevailing factor responsible for the formation of higher seismicity should be recognized the influence of interplate processes. Yet, this does not mean that it is horizontal movements that define the modern geodynamics in Fennoscandia. Computed seismic moments of earthquakes are too low to explain amplitudes of differentiated vertical movements established from repeated leveling (Slunga,1989). This leads to a logic conclusion on aseismicity of the most part of vertical motions. The same is suggested from the existence of vertical movements both on seismic and aseismic faults. At the same time, a fairly large amount of data evidence for both existence of local sites where other tectonic factors prevail in the creation of seismic stresses, and distorting, filtering effect on the observed stress field of fault tectonic in one or another particular region. These are data on focal mechanisms different from strike-slip, as well as a considerable scatter often observed in azimuths of subhorizontal compression axes.

Dense fragmentation of the Baltic Shield with systems of mutually orthogonal faults defines its small-block structure and related kinematic and dynamic non-uniformity of the motions. It is evident that in zones of the most intensive differentiated vertical motions a part of them will be realized as earthquakes. This can be demonstrated by following examples. The most obvious clusters of epicenters are observed in the area of such zones of maximum cumulative late post-glacial uplifts as the west coast of the central and northern Gulf of Bothnia (up to 250 m), Southwestern Sweden, Southwestern Norway, and the coast of the Northern Norway (local highs up to 150-200 m with a sharp gradient), northern Central Finland and Karelia near the Kandalaksha Bay (up to 200 m with a sharp gradient), the area of differentiated movements near Venern Lake (Figs.23, 24). In these terms, very indicative is rather high seismicity of the central part of the Kola Peninsula, where intensively uplifting Khibiny and Lovozero rock masses, within which one of the uplifting highs occurs (up to 200 m), alternate with the subsiding Lovozero and Umbozero depressions. At the same time, the entire eastern part of the Kola Peninsula suffering in general an insignificant uplifting, and presently beginning to subside (Nikonov,1980) is almost aseismic, except for the coastal zone of junction of the Baltic Shield and Barents Platform (Figs.15, 22). In the zones of maximum uplift it is natural to expect a different type of a focal mechanism, which has been supported by a solution on the Solberg earthquake dated September 29, 1983 (northwest coast of the Gulf of Bothnia), according to which a close to normal fault mechanism has been obtained (Kim and others,1985; Wahlstrom,1989) typical of the extension areas which are axial and off-axial sites of uplifting zones.

Fig.22 Map of earthquake epicenters in Fennoscandia

Fig.23 Chart of cumulative late post glacial uplifts in Fennoscandia

The influence of the features of the fault structure on the formation of the stress fields may be illustrated also by examples. So, with general similarity of stress fields of the Southern and Northern Sweden a much greater scattering in azimuths of horizontal compression axes from the general NW-SE trend in the North Sweden is noted. This may be attributed to the difference in the orientation of mutually normal fault systems in the aforesaid areas: meridional-latitudinal in the south and diagonal in the north (Slunga,1989). The influence of the local fault structure may account for the existence of focal mechanisms different from strike-slip one which is the most typical of Fennoscandia. It is clear that propagating from the northwest and north-northwest horizontal compression stresses relaxed primarily by strike-slip motions along the faults of the strike close to the above, on orthogonal faults should encourage the creation of thrust environment which may discharge as earthquakes. It is not accidental that at the bulk determinations of focal mechanisms from weak earthquakes, thrust solutions are the second frequent after the strike-slip (Slunga,1989). Taking into account the greater mechanic strength of the rocks at compression than at extension and shear (Spravochnik fizicheskikh..,1969) it should be expected that thrust motions will occur at strong earthquakes. This is what actually observed: with 8 available individual fault plane solutions by means of first motions method in 6 cases a thrust or close mechanism was obtained (Arvidsson and others,1991; Assinovskaya,1994; Avetisov and Vinnik,1995; Kulhanek and others,1981; Kvamme and Hansen,1989; Regional Catalogue). For 2 of them occurring in the Southern Sweden, according to geological data and distribution of the aftershock sequence, preferable fault planes coincident with faults of NE and NNE strike, orthogonal and suborthogonal to the orientation of the general compression, have been identified (Wahlstrom,1989).

The modern uplift is experienced also by rather seismically active areas such as Svalbard and Northeastern Greenland [6-7 mm/yr (Grigoriev and Musatov,1982; Semevsky,1967)]; there is information on earthquakes on Novaya Zemlya [4-5 mm/yr (Nikonov,1980)], Franz-Joseph Land [3-5 mm/yr (Kovaleva and others,1974)], Severnaya Zemlya [2-3 mm/yr (Nikonov,1980)] ). Therefore, combined analysis of the map of epicenters and materials on vertical movements of coasts and islands of the northern seas shows that all seismically active areas of the Norwegian -Greenland Basin and Eurasian Subbasin margins (as well as northeastern Arctic seas) are uplifting zones. At the same time, an opposite conclusion on seismicity of all uplifting zones cannot be confidently made now. Although weak seismicity of such presently uplifting areas as the Franz-Joseph Land, Novaya Zemlya, Severnaya Zemlya or aseismicity of Taymyr may be just apparent because of lack of observations, it is necessary to recognize the influence on this factor of the uplifting rate and consider the values of 5-6 mm/yr as a conterminous. A higher seismicity has not been reported from submerging areas.

In summary, the general conclusion is that higher seismicity of the margins of the Norwegian-Greenland Basin and Eurasian Subbasin is due, on the whole, to stresses transferred from the interplate zone of the mid-oceanic ridge and superimposed on the lithosphere activated by modern uplift. The observed seismic effect is a cumulative result of the operation of the aforesaid sources which directly depends on the distance of each particular region from the nearest segment of the interplate boundary, spreading rate along this segment, and uplift rate in the region. A similar seismicity level of Fennoscandia and Svalbard located at essentially different distances from the interplate boundary, more 500-600 km and 150-400 km, respectively, may be explained by the compensating effect of vertical movements of different intensity, 9-10 mm/yr and 5-6 mm/yr, respectively. Remaining margins, for which their distance from the boundary is nearly the same as that of Fennoscandia and uplift rates are commonly lower than Svalbard, are naturally characterized by much lower seismic activity.

Views of the reasons of modern uplift of the Norwegian-Greenland Basin and Eurasian Subbasin margins and, primarily, Fennoscandia and Svalbard have passed a way from common acknowledgment of its a purely glacioisostatic nature to presently prevailing opinion that glacioisostasy is subordinated while the leading role is played by deep tectonic processes. The glacioisostatic hypothesis was based on the scheme developed by De Geer-Hogbom in the beginning of the century (De Geer,1913; Hogbom,1913), according to which Fennoscandia has experienced dome-shaped uplift with the maximum rate in the center of the Baltic Shield, northern Gulf of Bothnia where the glacier was the thickest, and with rates decreasing from the center to the periphery. This opinion has been supported by calculations (Gutenberg,1941, etc.) showing that the ice sheet of 500 km in diameter and at least 1 km thick should cause downwarping of the earth crust. The following principal facts have been responsible for the viewpoint rejecting a certain role of glacioisostasy in the generation of modern vertical movements.

First, it is evident now, that dome-shaped uplift, if any, may be only suggested in the central part of the shield. The scheme of cumulative uplifts apart from the central high, shows periphery ones located outside the areas of maximum development of the ice cover. (Nikolaev,1966;1967;1988, etc.). As it has been mentioned above, distribution of these highs well correlates with that of epicentral clusters.

Second, numerous studies (Andrews,1970; Grachev and Dolukhanov,1970; Morner,1978; Quaternary geology.., 1989, etc.) have shown that maximum uplift rates generally occur not later than 1 thousand years after complete deglaciation and decrease in time under the exponential law. A drastic decrease in rates occur in 1-1.5 thousand years after the maximum. Over the first thousand years at least one third of uplift takes place. This is the period when a leading seismogenic part of glacioisostasy is possible. Thus, in the Northern Fennoscandia as it is shown in (Wahlstrom,1989) there is an evidence that a few of exposed faults of late - and post-glacial origin with vertical displacement over 20 m and some fault scarps were formed either simultaneously or within a short time (presumably a few decades) as a result of energy release at one or several strong earthquakes. The final isostatic compensation occurs according to common assessment not later than in 8-9 thousand years, while according to a proven opinion (Morner,1978) the effect of the glacioisostatic component on the uplift of the Baltic Shield ceased in 900 years. In any case, the fact of overall deglaciation in Fennoscandia not later than 9-9.5 thousand years ago allows a conclusion to be made that glacioisostatic movements in this region, probably save for the area of maximum isostatic load coinciding with the central part of the Bothnia-Kandalaksha depression at best do not play now any considerable part in the formation of seismogenic stresses.

Third, according to geological and paleogeographic data, H.Stille (Stille,1955) stated that Fennoscandia has experienced uplift since earlier epochs that that of the recent glaciation. Zero contour of the Baltic Shield presently is in the same position as it was during the Late Paleozoic (Late Permian), that is the modern uplift is inherited from the older one. Similar conclusions have been obtained on Svalbard and Greenland. Thus, according to D.V.Semevskiy (1967), the process of uplift of Spitsbergen was of different intensity during a long period of time commencing in the north of the archipelago in the Late Devonian. Lack of glacioisostatic component in Holocene (38-10 thousand years) uplift of the Svalbard and Greenland fjord coasts is proven by the existence there at that time of typical marine sedimentation (Grigoriev and Musatov,1982).

In our view, all available facts and opinions suggest that basically in areas, suffering glaciation vertical movements included both glacioisostatic and tectonic components. Their contribution to the cumulative motion considerably, from zero to maximum, changes depending on the location of the region and temporal period under study. The leading part of glacioisostasy may be recognized as it has been mentioned above during deglaciation and the first 1-3 thousand years afterwards. At present, glacioisostatic component is limited to the role of the regulator of vertical movements the effect of which yet ranges between the areas. Thus, for example, within Fennoscandia its role is more essential in the area of maximum glaciation (northeastern Gulf of Bothnia) and minimal, for example, within the confines of Norwegian Caledonids. Apparently, one should pay tribute to a considerable contribution of the glacioisostatic component to uplifting of coastal areas of Greenland which have been recently released from ice and show a significant seismicity. Aseismicity of inner areas of Greenland as well as Antarctic where contemporary tectonic uplifting is hindered by thick shields of modern glaciation also supports the idea that glaciation and deglaciation govern contemporary vertical movements.

Summing up the above, a conclusion may be drawn that the second strongest factor affecting the contemporary seismicity of the Norwegian-Greenland Basin and Eurasian Subbasin margins is not glaciation rebound but, in general sense, contemporary vertical movements of which presently non-crucial component is the glacioisostatic effect.

Finally, the most locally effective factor of the modern seismicity is the load of extremely thick sedimentary covers on the thin oceanic crust in the zone of its junction with the continental crust. According to calculations (Stein and others,1979), a sedimentary layer up to 10 km thick, which is quite realistic, may induce stresses up to 100 MPa (megapascal) and more, which is one order higher than those generated at discharge of interplate stresses and deglaciation. Basically, this level of stresses can ensure the highest seismicity along most Arctic margins characterized by thick sedimentary units, on the oceanic side in particular. Resolution of this obvious discrepancy between the theoretical estimations and observed seismic effect is like in many other cases in the properties of the real stressed lithosphere. The latter unlike the elastic layer capable of bearing immense stresses, begins destroying much earlier than maximum loading has been reached. As an alternative to the elastic medium viscoelastic and brittle/ductile models have been proposed (Stein and others,1979). The former has zero strength in respect of long-term loads, i.e. provides for a higher rate of aseismic relaxation (creep) as compared with their growth, and fairly high strength (up to 70-80% of the strength of the elastic layer) at fast growing stresses. It is evident that in this case the prevailing role in creation of excessive stresses is not played by the common thickness of the deposits but by changes in the thickness, i.e. intensity of modern sedimentation. Earthquakes are assumed to be realistic, for example, in the Gulf of Mexico (Nunn,1985) with sedimentation rate of 1.5 mm/yr. Within the margins of the Norwegian-Greenland Basin and Eurasian Subbasin an appreciable contribution of this factor to seismicity enhancement may be reasonably supposed for the Lofoten Basin, especially in the junction with the Norwegian and Barents Sea (Senja Fracture Zone) shelves. Here, a drastic increase of sedimentation rate was identified to be 1.4 mm/yr and 3-7 mm/yr in the Pleistocene and during the last 100 thousand years, respectively. (The Arctic Ocean Region..,1990). However, the second model seems to be more realistic, both in terms of common understanding of the environment and based on data available on fairly low seismicity of the intraplate parts of oceanic basins. In the upper part it is characterized by the pressure-induced growth of brittle strength with depth. However, due to high temperatures, material behavior in the lower part is controlled by ductile strength decreasing with depth. Evidently; such a medium precludes accumulation of high stresses regardless of their growth rate, which agrees well with the observed low seismicity level. Apart from the Lofoten Basin, seismogenic effect of the overburden load is also suggested in the Lincoln Sea. (Basham and others,1977; The Arctic Ocean Region..,1990).

Closing up the review of geodynamics of seismically active marginal zones of the Norwegian-Greenland Basin and Eurasian Subbasin, one can draw a justified conclusion that none of the above seismic factors is severally able to ensure high seismicity level. The observed seismic effect is the result of cumulative effect of at least two of them where compulsory is the presence of stresses generated in interplate zones. In view of possible fluctuations of absolute values of accumulated stresses intensified by inhomogeneity of the lithosphere one can also suggest that in each particular region or even case the role of primary or secondary factors will be alternately played by either of them. The strongest earthquakes are possible to occur in favorable (phase coincident) superimposed actions of all the three seismic sources.

The Arctic Canada and adjacent areas.

Comparison between the map of epicenters with the geological chart (Figs. 19, 20) suggests two, in our view, hardly arguable conclusions.

Firstly, the comparison indicates obvious tending of seismically active zones to contacts of blocks substantially differing in their geological development. There is no doubt that such contacts, along with disjunctive dislocations, are weakened lithospheric zones.

Primarily, one should note that epicenters distinctly trace almost the entire northwestern near-coastal boundary of the sedimentary Sverdrup Basin. Disappearance of the epicenters is only observed on its northeastern termination. The second seismically active zone extending from the continent in submeridional direction is undoubtedly correlated with the linear, expended far northward structure of the crystalline basement of the Boothia Uplift and its extension, as inferred from geological data, under depositional strata of the Franklinian geosyncline and Sverdrup Basin. In this case, one may say that distribution of earthquakes has provided independent information on the relief of the buried part of the basement. A characteristic feature is coincidence of two local epicentral swarms identified in this seismically active zone in Barrow and Bayam-Martin straits with its intersection of southern margins of the Franklin geosyncline and Sverdrup Basin, respectively. It is also interesting that a local oceanic zone of considerably higher seismicity lies approximately in the northern range line with this band of epicenters. Assumption of possible intersection of the buried structure of the Boothia Uplift and the continental slope (which is the contact zone between geoblocks of continental and oceanic types of crust ) may account for higher seismicity of this particular site of the continental slope with adjacent part being aseismic.

In this context the seismically active zones of the Baffin Bay and adjacent areas tracing marginal and contact sites of the Canadian and Greenland crystalline shields and also the junction of the blocks with the continental and oceanic lithosphere are not an exception. In the oceanic part of the bay, the role of weakened zone emplaced during the spreading stage of the basin evolution seems to be essential. In particular, one may share the opinion expressed by (Kroeger,1987; Quinlan,1984; Stein and others,1979) regarding the fact that the strongest earthquake of 1933 and its intense aftershocks are located in the zone of an ancient transform fault. The chains of epicenters in the Lancaster and Johns straits are likely related to similar weakened zones.

The second conclusion is that while all epicenters tend towards any weakened zones of the lithosphere, then in turn, not all obviously weakened zones are seismic. Thus, for instance, the eastern junction of the Sverdrup Basin with the adjacent zone of the Franklinian geosyncline, contact zones of the Banks Basin, Canadian crystalline shield west of the Boothia Uplift and its narrow long projection near Victoria Island (Minto Uplift) and others are aseismic. Frequently, even in seismic areas, without visible reasons, seemingly within the single zone sharp discontinuities and gaps in the distribution of epicenters are observed: on the east coast of the Baffin Island near 70° N between two thick swarms of epicenters, as well as in the northern and southern parts of its coast, in the central part of the Boothia Uplift, etc. This second conclusion raises a question of the presence of additional factors affecting deviations in the seismicity level. These are undoubtedly still unknown details of the deep structure of the lithosphere and particular features of stress source(s).

In the Arctic Canada, among three above factors responsible for the intraplate seismicity of the Arctic, the first one in view of the distance of thousands kilometers away from plate boundaries, in general should act similarly in any part of the region, being a kind of a background. The response of the lithosphere to release of the glaciation load should be recognized as a regional factor because the last glaciation covered nearly the entire study area. However, its role must be dramatically variable in different points and depend on the thickness of the glacier and deglaciation time. Finally, the third of the aforesaid factors undoubtedly has a local effect, and its part may be only essential in the Baffin Bay and Arctic Ocean. Different stress combinations superimposed on the heterogeneous and inhomogeneous geological structure often poorly studied, create to the observed complex pattern of seismicity in the above region. This complicates evaluation of the contribution of each factor in a specific area, thus precluding any conclusion on common regional correlations to be drawn.

Acknowledging in general the presently existing assessment of the tectonic nature of earthquakes in the Arctic Canada one should note that available geological, geophysical and glaciological data suggest some additional ideas regarding the role played by certain of the aforesaid factors.

In our view, it is impossible to share the opinion of (Basham and others,1977; Stein and others,1979; Wetmiller and Forsyth,1982) in respect of good seismicity correlation in the entire Northern Canada with areas of maximum differentiated glacioisostatic movements.

The presence of glacioisostatic movements in the Northern Canada, as in Fennoscandia, was identified in the end of last century. Higher seismicity of both regions, which was later instrumentally recorded, had been attributed by all scientists to deglaciation However, no special attention had been paid to the evident fact that, in each region, higher seismicity zones were distributed quite differently with respect to areas of different degree of glaciation, and hence of different intensity and differentiation of glacioisostatic movements. While in Fennoscandia, the coincidence of the zone of maximum glaciation and naturally maximum glacioisostatic movements in the central part of the Baltic Shield (northern part of the Gulf of Bothnia) with the zone of maximum seismicity can be assumed, then in the Arctic Canada, zones of maximum glaciation and maximum movements near the Fox Basin and on northeastern of the Ellesmere Land (Fig. 24) are almost aseismic like the thickest zones near the Hudson Bay (Andrews,1970; Quaternary geology..,1989; Wetmiller and Forsyth,1982). Higher seismicity tends rather towards the periphery of the glaciation shield (Fig.19). This circumstance alone, in view of similarity of tectonic settings of the above regions suggests that either in one of them or in both glacioisostasy does not play crucial part in modern higher seismicity.

As it has been mentioned above, at present it may be taken for granted that nearly entire compensation of the glacial load release occurs under common estimation within 8-9 thousand years. This has been the main argument in favor of rejecting the leading part of glacioisostasy in the modern vertical movements in Fennoscandia.

For the Arctic Canada arguments using this estimation seem to be less convincing because overall deglaciation took place here later, namely 6-6.5 thousand years ago. However, it should be kept in mind that these figures pertain to the centers of maximum glaciation. At the same time, seismic northern part of the Canada Arctic Archipelago, except for islands, such as Ellesmere Land and Axel Heiberg, did not experience recent glaciation and did not suffer during this period any vertical movements (Quaternary geology..,1989). By 9-9.5 thousand years ago apart from presently seismic Baffin Island and aseismic Ellesmere Land, all the remaining areas of the archipelago both seismic and, released from glaciation. Therefore, the absence of correlation between the distribution of modern seismicity and the time of deglaciation, in our view, suggests that in the Arctic Canada glacioisostatic movements at present in general do not play essential part in the formation of seismogenic stress field.This conclusion should not be extended to the west coast of Greenland which is just being involved in the deglaciation process.

In respect of glacioisostasy assessment today the map of modern uplifting rates presented in (Andrews,1970) is interesting. It shows that the zones of the highest rates not exceeding 8-13 mm/yr are located in different sites than 6-8 thousand years ago when glacioisostatic movements prevailed. These are, primarily, the northeast mountainous coast of the Baffin Island where concurrently submergence of coastal beaches was identified, and less expressly, seismic northern part of the Canadian Arctic Archipelago and, beyond the scope of the study area, seismic southeastern part of the Labrador Peninsula (Fig.25).

Additional information generally supporting the presented conclusion on the role played by glacioisostatic movements in the formation of the modern seismicity field provides an analysis of focal mechanisms. In (Talwani and Rajendran,1991) based on data obtained from 29 strong earthquakes which occurred in various sites

Fig. 24 Mean rates of the post glacial uplift in northern and eastern North America

Fig. 25 Recent rates of the post glacial uplift in northern and eastern North America

of the globe, two major types of intraplate earthquakes have been identified by particular focal mechanisms: A-type - strike-slip movements with steady in most regions orientation of stress axes caused by tectonic movements on plate boundaries, and B-type - other models of motions arising from local disturbing stresses superposed on the background A-type field. As is mentioned above, it is strike-slip mechanisms that are observed in most part of the study region. They show sufficiently steady parameters and compression axis nearly orthogonal to the boundary of the North American Plate in the Eurasian Subbasin of the Arctic Ocean (Fig.20). This fact suggests that the background stress field is poorly distorted here by local disturbing objects, and these distortions raising the overall seismicity level are not strong enough to change the type of the focal mechanism. This is also supported, in our view, by fairly low intensity of earthquakes, of which the strongest had magnitude not more than 5.5, and by the fact of fairly quiet gravity field (Sobczak and Halpenny,1990) in the areas of A-type earthquakes, which definitely evidences for predominant role of horizontal motions in comparison with vertical ones. Logically, it should be assumed that background stress alone, without other disturbing factors involved, is insufficient for occurrence of earthquakes, which explains the existence of aseismic areas.

Anticipated exceptions are the earthquakes of the Baffin Island and Baffin Bay where fault plane solutions in strict compliance with the data presented in (Andrews,1970) and the map of isostatic anomalies (Sobczak and Halpenny,1990) are an evidence of uplift of the shield edge and submergence of the bay bottom. Land uplift, as may be assumed from the above, is caused not by glacioisostatic causes, but likely, by epeirogenic movements of deep nature. The same mechanism underlies the bay bottom submergence, which is well fixed by a distinct intensive positive isostatic anomaly on the gravity map. The existence of the above movements alone provides favorable conditions for earthquakes and, in the long run, everything depends on the quantitative ratio between the value of accumulated stresses and strength peculiarities of the lithosphere. Just as regional tectonic movements superimposed on the background intraplate stress field distort it, factors of lower order are able to attenuate or enhance the effect of these movements, which ultimately can and does give rise to very strong earthquakes of magnitudes 6 and more. Within the Baffin Bay such reasons, apart from already mentioned non-revealed irregularities of the deep structure may be the load of extremely thick sedimentary cover or rather, as it has been shown earlier, abnormally high sedimentation rate. They may lead to local raising of amplitudes and differentiation of movements of the lithosphere, and occurrence of earthquakes. It is known that the thickness of sedimentary cover in the oceanic part of the Baffin Bay reaches 8 and more kilometers, while sedimentation rate drastically increased from 0.1-02 mm/yr in the Neogene to the highest in the Arctic striking value of 15 mm/yr in the Holocene (Geology of Greenland,1976; Nassichuk,1983). Most likely that the role of this factor is essential also in the seismically active zone of the Arctic Ocean adjacent to the Canadian Arctic Archipelago [thickness of the cover is up to 12-15 km (The Arctic Ocean Region..,1990), data on sedimentation rate not available] and oceanic part of the Beaufort Sea [thickness of the cover up to 10 km (The Arctic Ocean Region..,1990), sedimentation rate in the Holocene 2.5-2.7 mm/yr (Nikonov,1980)] where the presence of sharp gravity step is an evidence of the predominant part played by vertical movements.

It should be noted in general, that awareness of the results of seismological studies in the Arctic Canada has shown underestimation by seismologists of vertical (non-glacioisostatic) movements as a factor capable of leading to earthquakes, although in geoscientific literature there are many references to the existence of such movements throughout the history of the region. The aforementioned map of modern tectonic movements (Andrews,1970) and data on multiple vertical motions on the Boothia Uplift and Peary geosyncline along a number of fringing faults, and on three stages of impulsive submergence of the Sverdrup Basin (Sweeney,1976;1977) may be assigned to such information. The conclusion on the existence of differentiated vertical tectonic movements between Prince of Wales Island and Boothia Peninsula, and Somerset Island was made by Quaternary geology(1989) on the basis of a definite (100-110 m) non-coincidence in the positions of shorelines of the same age.

Comparison we have made between plots of energy annually released by earthquakes in different zones of the Arctic Canada and mid-oceanic ridge zones in the Norwegian-Greenland Basin and Eurasian Subbasin of the Arctic Ocean supports the idea that none of aforesaid seismic factors alone may not be responsible for emergency of seismically active zones. For the whole period of instrumental observation (since 1909), earthquakes with magnitude 5 and over, which were used for comparison, have turned out to be representative. This comparison has shown that there is no notable correlation between seismic regimes of the aforesaid interplate zones and intraplate seismicity of the Arctic Canada either on the whole or by individual area. The cross-correlation coefficient in different time shifts did not exceed 0.2-0.3. It should be noted that we obtained a similar result for earthquakes of the Beaufort Sea.

For comparison between the regimes of seismically active zones within the Arctic Canada the time period was taken between 1960 and 1990 during which earthquakes with magnitude 4 and over were representative. Correlation in time has not been established here either, even for neighboring zones, e.g. the Baffin Island and Baffin Bay. The only exception is the aforementioned coincidence of bursts of seismic activity in 1972 in the Bayam-Martin Strait and in the zone northwest of the archipelago. Besides, noteworthy is the obvious common recession in seismic activity in the Arctic Canada which began approximately after 1975 and lasted at least until 1990.

So, summing up the above, the following conclusions may be drawn.

The present day seismological data indicate the existence in the Arctic Canada and adjacent areas of seismically active zones localized rather well. All of them trace the contacts of blocks with different tectonic evolutionary history confirming the fact that earthquakes tend towards weakened zones of the lithosphere. At the same time, a number of similar contacts is aseismic.

The higher seismicity is caused in general by cumulative effect of a few tectonic factors, among which background intraplate stresses emerging as a result of tectonic processes on lithospheric plate boundaries are common of the entire region. Other factors are vertical tectonic movements and, in offshore areas, local effects of higher thickness of the sedimentary cover and sedimentation rate. As it is shown by the analysis of glaciological and geological-geophysical data the contribution of glacioisostatic movements in the modern tectonics and seismicity of the Arctic Canada and adjacent areas, likely except for the west coast of Greenland, is negligible.

Areas with prevailing background stresses are characterized by strike-slip motions, fairly quiet gravity field and relatively low seismicity level. In the zones of background stresses only singular and weak earthquakes may occur. In the areas of joint effect of several factors other types of focal mechanisms, drastically variable gravity field and higher intensity of earthquakes are observed. The strongest earthquakes occur in weakened zones of the lithosphere with coherent (resonance) adding of all seismic factors. Individually, each of them, background stress, in particular, is not capable of causing high seismic activity. An indirect support of this is the absence of correlation between temporal characteristics of regimes in intraplate seismically active zones of the Arctic Canada and interplate Norwegian-Greenland Basin and Eurasian Subbasin of the Arctic Ocean.

Closing up the review of the modern geodynamics of seismically active zones of the Arctic Canada one cannot avoid brief discussion of the problem of the position of the Mesozoic boundary between the North American and Greenland plates. According to plate tectonic reconstructions based on the interpretation of magnetic data, and as it has been shown above, this boundary existed during the period between anomalies 34 and 13 (70-36 million years; the Late Cretaceous-Early Oligocene) and was marked from the North Atlantic through the Labrador Sea, Davis Strait, Baffin Bay and farther to the Nares Strait dividing Greenland and the Canadian Arctic Archipelago, north of which beginning anomaly 25 it formed a triple junction with interplate boundaries of the Norwegian-Greenland Basin and Eurasian Subbasin (Jackson,1985; Kristofferson and others,1977; Pitman and Talwani,1972; Srivastava,1978;1985, etc.). While the location, kinematics and dynamics of most of the boundary on the whole have been quietly accepted, the northernmost part thereof has aroused a hot discussion. The problem is in the dramatic contradiction between quantitative parameters of motions along the boundary obtained from plate tectonic reconstructions, and data from onland geological studies on either side of the strait. According to the former, along the strait there was a sinistral strike-slip with the cumulative amplitude of 125 km (Srivastava,1985) or even 250 km (Jackson,1985), and compression of 90 km (Srivastava,1985) or 120 km (Jackson,1985). Data of the vast majority of geologists dealing with this region based on tracing in the strait of numerous distinct markers, and analyses of the Eureka Folded Belt do not accept the strike-slip and compression over 25 km and 40 km, respectively (Christie and others,1981; Dawes and Kerr,1982; Higgins and Soper,1989; Kerr,1967;1981; Okulitch and others,1990, etc.). The conclusion regarding breaks in continental crust under the Nares Strait was drawn in (Wetmiller and Forsyth,1982) based on the study of Lg waves which are known not to propagate in the areas with the oceanic type of the crust.

Within the scope of the present work dedicated to the modern geodynamics of seismically active zones of the Arctic Region we would like just to present some ideas regarding this matter.

Primarily, in our view, it should be postulated that despite striking similarity of geological settings on either side of the Nares Strait it is evidently a fault structure. To assert this, a simple acquaintance with geomorphological characteristics of the region is sufficient: elongation of the strait (about 600 km), small width (mainly, 40-60 km), great bottom depths (450-500 m), steep mountainous coast with absolute marks up to 900-1000 m. Deep protrusion of the strait in the ancient units also including the Archean basement in the southern part of the Canadian Shield, should not challenge the deep character of this fault.

It is clear that such enormous faults both in lateral and depths dimensions or lineaments, basically may be the arena of tectonic movements with those cumulative amplitudes which are proven by plate tectonic reconstructions. In this connection, it should be noted that aforementioned estimations of the lithosphere of the Nares Strait area made on the basis of analysis of Lg wave propagation, in this case are not important in the settlement of the dispute between proponents and opponents of great movements, because neither of them does not prove the presence under the strait of oceanic crustal blocks.

Yet, despite of the recognition of basic possibility of large scale motions along the Nares Strait, in our view, the preference in this, like in other similar cases, should be given to conclusions made from the results of onshore geological studies. The example in question is fairly typical and instructive, and again disclose, in our opinion, two bottle-necks in the position of orthodox mobilists in their long dispute with opponents of large-scaled horizontal movements.

The first one is in the effort to solve just regional or even local geological problems applying methods and approaches allowing but transregional or global assessments to be obtained.The author, for example, has been always impressed with the miraculous capability by some scientists of confidently identifying coupled anomalies, assign numbers to them and trace them along hundreds or even thousands kilometers, compute coordinates of poles of opening located dozens thousands kilometers away, and as a result of all these fairly approximate reconstructions to present sizes of motions in the local area of the amplitude of dozens kilometers with an accuracy of 5 to 10 kilometers, using a complex magnetic field where even simple linearity of anomalies is hardy traced. It is similar, for instance to attempts to solve problems on differentiation of finely laminated medium based on the observations using low frequency seismic instruments recording waves of several kilometers long. This is exactly the situation with the problem of the Nares Strait. Of course, adoption of the Nares Strait as an extinct lithospheric plate boundary seems to be very attractive and logic. Legitimate appears the attempts to compute the amplitude and the type of tectonic movements in terms of plate tectonics, however, the results obtained should only have been regarded as the very first approximation and subject, in our view, to immediate revision after collection by onshore geological studies of numerous reliable data refuting these assessments.

The second of the aforesaid bottle-necks in mobilists position also highlighted by the problem of the Nares Strait implies a blind, unconditional pursuance to the principle of "absolute plate rigidity" and "persistence" of interplate boundaries which are virtually needed and sufficient but on a global level for illustration of basic terms of plate tectonics. In transition to larger scale, it is necessary to take into consideration the fact that a significant part of stresses may relax by intraplate movements, as a result of which, the scale of movements on interplate boundaries should be reduced. Orthodox mobilists ignore the facts of the existence of large-scale folded belts emplaced in different times, the wealth of data regarding intraplate seismicity, information on the entire system of modern continental rift zones not yet connected with any contemporary lithospheric plate boundaries - i.e. irrefutable geological events evidencing that not all tectonic stresses discharge on interplate boundaries.

Different ideas are likely to be expressed on the dynamics of the lithosphere in this region during the time period between anomalies 34 and 13, where data on onshore geological studies and the fact of location of the Greenland plate boundary along the Nares Strait will be taken as the basis. The common point of all these ideas should be that of a greater differentiation of the Greenland plate and adjacent areas, which result in stress distribution and accordingly, stress-induced motions along several weakened zones, thereby lifting the question of assigning a cumulative motion to any of them (Nares Strait). The difference will be in assumed locations separating the blocks of weakened zones. One of these assumptions implies, for instance that the zone of extension of the Baffin Bay bifurcated in its northern part and kept on extending not only northeastward near the Nares Strait, but also westward to the Lancaster Strait (Fig.19). A broad area for similar assumptions is Greenland due poor knowledge of this part where the existence of weakened zones splitting it into minor blocks may be expected. It is known that during the Paleogene, the North Atlantic experienced basalt volcanism, immense both in terms of intensity and propagation area. It occurred likely in the environment of considerable split of the lithosphere. It is supposed that a Transgreenland deep fracture zone formed at that time along 70° N (Milanovsky,1976;1977). Taking into account of such zone may radically change the character of lithospheric block movements in this region, and debase plate tectonic reconstructions presented by (Kristofferson and Talwani,1977; Srivastava,1985).

The closest to the author¸s is the viewpoint by J.Kerr, first expressed in 1967 (1967; 1977, etc.) and based on the idea, which natural for impartial scientific opinion, that a graben-like structure of the Nares Strait is an extension zone which is measured by the width of the strait rather than by the strike-slip or, moreover, compression zone. According to J.Kerr, the Labrador Sea, Davis Strait, Baffin Bay and Nares Strait are an polygonal line system of extension zones cross-cut with faults. Its amplitude is decreasing northward reaching its minimum in the Nares Strait. The width of extension zones south of the Nares Strait is smaller than required in computations by (Jackson,1985; Srivastava,1978;1985) from 1000 m contour. The existence of oceanic crustal blocks beyond extension zones may be accounted for their pre-spreading formation.

The aforementioned detailed geological studies have shown that the location of the Eureka folding front on the Ellesmere side of the strait is 200-250 km farther south than on Greenland side. This has been the main argument of proponents of considerable strike-slip along the strait. Actually it is attributed to the lateral bending of this front northward, which has been caused by particular development of an older (Middle Paleozoic) Ellesmere Folding (its missing from the Greenland side of the strait). This, in turn, is reflected from the differences in the basement relief on either side of the strait, particularly, its arch-shaped rise on the Greenland side of the strait hindering southerly propagation of folding fronts. Therefore, the contemporary environment of the Nares Strait area is closely related to the history of its previous evolution, while the strait itself is an ancient fault structure which had been reactivated from time to time.

The near-Pacific Arctic Region

The main reason for special brief discussion of the seismicity in the aforesaid region primarily consisting of the North Alaska, Chukchi Peninsula and adjacent northern offshore areas is the geographic proximity of these Arctic zones to the most worldly active Pacific Seismic Belt tracing the convergent boundary of the lithospheric plates, which is antipode to divergent type of boundaries conjugated with mid-oceanic ridges. This fact poses a natural question whether there are any peculiarities in the dynamics of the lithosphere of the near-Pacific Arctic areas making them different from the above mentioned seismically active zones. Basically, in this case, the situation is that primarily the question should be posed in general, what type, either interplate or intraplate, the seismicity of the above pacific zones should be assigned to. What is the relation between the generally sublinear seismically active zones identified here and the general Pacific Fire Ring Zone? Legitimacy of this question is confirmed by the opinion, for instance, of A. Grantz et al. (The Arctic Ocean Region..,1990) that the Canning Displacement Zone of is the northern extension of the Aleutian Benioff Zone from the Central Alaska to the Beaufort Sea shelf, and the tracing band of epicenters is the northern fragment of the global pacific belt of interplate seismicity. This statement can be logically continued by assignment to this seismicity type of the Brooks Range Zone the band of epicenters of which merges with that of the Canning Zone.

The entire structure of this paper shows that the author has an opposite view. The basis for this view was consideration of basic parameters characterizing and identifying each type of seismicity in general, and in given regions, in particular. These parameters include geographical distribution of earthquakes, depth of hypocenters, focal mechanisms and seismic activity level.

Fig.6 shows that merger of near-Pacific Arctic seismically active zones and the proximal Aleutian-Alaskan Zone of interplate seismicity can be hardly assumed. It is apparent that the easternmost one, extended within the Northwestern Territories along Mackenzie Valley and Mackenzie Mountains is isolated. Moreover, this zone as it was mentioned above, consists of nearly disintegrated separate links. The Brooks Range and Canning Displacement Zones do not merge with the Aleutian Alaskan. They rather join it on a small site near 148° W moving apart from each other to hundreds kilometers westward and eastward from this site. Blind terminations of the seismically active zone "the Brooks Range-Canning Displacement Zone" are seen in the west along the Bering Strait coast, in the north off the Beaufort Sea, and in the east near 138-140° W.

Fairly remarkable distinctions between the above zones has been shown by comparison of depths of hypocenters summarized in Table 43. In the comparison used the aforesaid Arctic seismological data bank.

Table 43

Distribution of depths of hypocenters and earthquake magnitudes

N - number of earthquakes with M > 5 per 100 thousand sq. km ("specific" activity )

* - one of the earthquakes has magnitude more than 8 (8.3)

The table presents a total number of earthquakes with defined depths of hypocenters recorded for each region including percentage of events at depths deeper than 50, 100 and 150 km, respectively. The table expressly shows a considerably deeper penetration into seismic focal zone of the South Alaska which is a northern fragment of the Pacific interplate seismic belt. Nearly all earthquakes of the seismically active zones of the Arctic Canada and Fennoscandia,which are undoubtedly intraplate by nature, are shallow. Similar data have been obtained from the Northwestern Territories and interplate divergent boundary of the Gakkel Ridge. Rejecting absolutely incredible assumption that the seismically active zone of the Northwestern Territories is part of the divergent plate boundary one can suggest that this is another evidence of its intraplate character. Fairly larger than in typical intraplate zones percentage of earthquakes with the depth of hypocenters ranging between 50 and 100 km is reported from the Brooks Range and Canning Displacement Zones. However, even in this case none unbiased researcher would suggest the existence of deep seismic focal area like the Benioff Zone. Additional confirmation to the above is the fact that there is no relationship between distribution of hypocenters laterally and vertically as it takes place in the Benioff Zone: earthquakes with deeper hypocenters are scattered throughout the entire area of seismically active zones.

As it has been shown above, in the Pacific Arctic seismically active zones there is no regular arrangement in focal mechanisms common of any of the two interplate boundaries. With the generally significant role of subhorizontal compressing stresses and trust motions a larger spread of stress axis strike azimuths, as well as equally frequent as trust, strike-slip and even normal fault motions (Fig.17). This is an evidence of the presence likely of more than one stress source, as well as of the complicating influence of the geological environment without a fracture zone prevailing any other (like Benioff Zone) and capable of regularly arranging the observed stress field.

Finally, it is evident, that compared to the Pacific Belt, the adjacent Arctic zones show much lower level of seismic activity. This is qualitatively seen from Figs.6 and 17 and quantitatively confirmed by data from Table 43. In order to avoid distorting effect of the aforesaid irregularity in the Arctic stations coverage, the table shows for each region a number of earthquakes with magnitude 5 and more (6 and 7 inclusive) which had exceeded completeness threshold in the Arctic throughout the years. In addition, the number of these earthquakes is reduced to the unit area of

100, 000 square kilometers.

The table shows that the opposite poles of the seismic activity levels with a difference of one and a half to two orders, there are the highest activity on the convergent plate boundary and lower one common of intraplate zones of the Arctic Canada and Fennoscandia. Seismicity of the divergent interplate boundary ranks second lagging far behind. The lagging becomes essential for the strongest earthquakes where the maximum of seismic energy is released. The intermediary position between interplate and intraplate zones being expressly different from either of them are occupied by near-Pacific Arctic zones.

Therefore, the discussion of the above seismic features seems to lead to a dead-end: regimes of near-Pacific Arctic seismically active zones show a number of features preventing from accepting their genetic unity with the immediately proximal interplate zone of the Pacific Seismic Belt. At the same time, they have essential distinctions from the aforesaid intraplate zone of Fennoscandia and Arctic Canada, especially in the depth of hypocenters and level of activity. However, deeper insight into the problem make it clear that the dead-end is just apparent. As it said in the discussion of the intraplate seismic zones, the regional factor frequently the most significant one responsible for particular features of each given zone is the influence of the interplate seismicity zone most closely located to it. This influence governs not only the level of the intraplate seismicity, but also focal mechanisms features. Therefore, it is natural to expect notable variations in parameters of intraplate zones which are in a way secondary subject to particular features of primary interplate zones. The closest to Fennoscandia and the Arctic Canada is the Mid-Oceanic Ridges Belt which is the least active among the three global seismic belts of the earth; besides, the Arctic Canada is located at a fairly great distance away. At the same time, near-Pacific Arctic regions are in immediate proximity or even adjoin to the strongest Pacific Seismic Belt, and in our view, owing to these geographical differences, their seismicity showing basic intraplate character has notable differences in individual parameters. As it has been shown above, the range of these parameters, especially of intensity level, may drastically increase due to other superimposed seismogenic factors and in "resonance" cases to result in the strongest earthquakes like in the Baffin Bay dated 1933. For the near-Pacific Arctic Region, two examples of the strongest damaging earthquakes may be provided: these occurred nearly in one point of the Northwestern Territories in October-December 1985 with magnitudes 6.5-6.7 and by certain evaluations, 7-7.2, and hundreds foreshocks and aftershocks with magnitudes up to 5 and more.

According to geological data (The Arctic Ocean Region..,1990) the folded structure of the arc-shaped system "The Brooks Range - Richardson Mountains - Mackenzie Mountains" began forming in the Middle Jurassic as a response to convergence between the paleobasin of the Pacific Ocean and southern margin of the Arctic Alaska. This process was completed in the Middle Early Cretaceous (Albian Stage) and with this the existence of the systems as an active convergent plate boundary has, in our view, terminated. Most likely, the Eocene trust formation established in the eastern part of the Brooks Range was affected by stresses generated in the new zone located south of the plate boundary. Therefore, near-Pacific Arctic zones associated to the outer (from the Pacific Ocean) line of the North American Cordillera may be regarded as the Late Mesozoic to Early Paleozoic interplate boundary moved far back. This is presently an arena of intensive intraplate seismicity caused by its proximity to the modern boundary of the North American and Pacific plates. Tectonic activity of the zone is enhanced by differentiated vertical movements mentioned above, though it is hard to identify the share of these movement which is not affected by horizontal stresses generated on the closest plate boundary.

Unconditionally intraplate characteris manifested by seismicity of the entire Northeastern Russia, including the Chukchi Peninsula. The location of this wedge of the North American Plate between the divergent (in the west) and convergent (in the east) boundaries is responsible for the formation here of a complex stress field investigation and understanding of which is impossible without special detailed seismic observations.